Code & AI agent sandboxes: read the guide

Agent Sandbox

Open source toolSandbox

agent-sandbox enables easy management of isolated, stateful, singleton workloads, ideal for use cases like AI agent runtimes and reinforcement learning (RL).

Homepage Docs Repository

Scanned 8 Oct 2026 · Schema v5 · features v21

Licence: Apache-2.0
Self-hosted; you pay for the infrastructure.
Free tier: Free to run
Isolation: Kata / gVisor
hardened or user-space kernel
Regions: —
Not disclosed on the public surface.
Capabilities: 14 / 61
11 partial · 32 fail · 0 not assessed
Agent readiness: 7/10 · 8/9
read · drive, pre-login surface

Advertised for

AI agent runtimes · Reinforcement learning · Stateful singleton workloads

Built for

Developers

· Kubernetes CRDs, kubectl, cluster installation, Python/Go SDKs; operator-configured isolation and network policies.

01 Runtime Each fact carries its source; dots show confidence. Open a row for the evidence.

Isolation & lifecycle

Isolation primitive

Kata / gVisorhardened or user-space kernel

Persistence

Kubernetes persistent volume claims

Snapshots / forks

Conditional: GKE PodSnapshot memory and filesystem restore

Resources & limits

Limits & quotas

Cluster and namespace resource limits; no vendor quota

Access

Deploy interface

cliapisdk

SDK languages

PythonGoTypeScript

Network & secrets

Egress controls

Operator-configured NetworkPolicy and enforcing CNI

Secrets

Kubernetes Secrets

Bring your own cloud

Placement & use

Jurisdiction

Advertised workloads

AI agent runtimesReinforcement learningStateful singleton workloads

02 Capabilities 14 pass · 11 partial · 32 fail · 61 of 61 assessed

Supported — Documented on the public surface and verified.
Partially supported — Exists with caveats: gated, thin, or enterprise-only.
Not supported — Checked and not offered.
Not assessed — Scanned before this capability was tracked. Not a lack.

Compute & storage 3/11

Persistent diskStorage that survives restart/stop (not ephemeral-only).

Supported

Object storageManaged S3-compatible / blob object store the platform PROVIDES to a user's workloads (buckets, presigned URLs). DISTINCT from persistent-disk (block/volume attached to an instance) and from a self-host dependency (source.self_host.deps[]=["object-storage"] is an infra dep to RUN an OSS tool — the opposite direction). SCOPE varies and matters: native first-class (Railway Buckets, Cloudflare R2) vs a sibling SKU / extension (Fly Tigris, Bunny Storage, DigitalOcean Spaces) — record which in the evidence quote / note. Admitted v15.

Not supported

Snapshots / forkingSnapshot or copy-on-write fork of a running instance (filesystem-only counts). Whether the snapshot also restores memory + running processes is recorded separately by memory-snapshot (v19).

Partially supported

Live-state (memory) snapshotsA snapshot / checkpoint / pause captures a running instance's MEMORY and PROCESS state and restores it, so running processes continue where they left off instead of a freshly booted instance with the same files. Counts whether user-triggered (snapshot, checkpoint, pause-to-storage) or used to restore pre-initialised state (memory snapshots for fast cold starts). Distinct from snapshots-forks (any snapshot or fork, including filesystem-only; a product passing that key may still fail this one) and from scale-to-zero (an idle-cost property; suspension may or may not keep memory). Actor multiplexing (see scale-to-zero) records its live-state half here — only when the suspended actor's MEMORY/process state is what comes back, not a freshly booted worker with the actor's files.

Partially supported

Scale to zeroIdle instances suspend at no / negligible cost. ACTOR MULTIPLEXING (many suspended stateful actors sharing a small pool of pre-warmed workers, woken on request — Agent Substrate, AX, kagent; Rivet's hibernating actors) has NO key of its own (proposal REJECTED v20) — score the idle-cost half HERE and the live-state half under memory-snapshot; the oversubscription ratio / density is a performance characteristic the registry does not key.

Partially supported

GPU computeGPU-backed instances available.

Not supported

Multi-regionCompute placeable in more than one geographic region.

Not supported

Sandbox templates / custom imagesBuild reusable sandbox/VM images from an OCI image or Dockerfile (a base you fork instances from).

Supported

Cloud-init / ignition configCustomise a VM / microVM / instance at boot via cloud-init (or ignition) user-data — the same boot-metadata interface as AWS/Azure/GCP. Inject packages, users, SSH keys, files, and run-commands declaratively at launch. Distinct from sandbox-templates (a PREBUILT image you fork instances from) — this is PER-LAUNCH declarative customisation, not a baked image.

Not supported

Desktop / computer-use sandboxA virtual desktop (GUI) for LLM computer-use — browser + desktop automation, often via noVNC/VNC.

Partially supported

Selectable isolation runtimeThe user chooses the isolation runtime per workload / environment / sandbox from more than one option (e.g. microVM vs gVisor vs plain container, native namespaces vs gVisor, gVisor vs VM). Records the CHOICE; the isolation_primitive facet still records the best-evidenced primary. A platform that switches runtime automatically without a user choice (e.g. gVisor forced for GPU) does not pass on that alone.

Supported

Networking & delivery 2/9

Automatic HTTPSTLS endpoints provisioned automatically per workload.

Not supported

Built-in reverse proxyPlatform routes external traffic to workload services.

Supported

Custom domainsBring-your-own domain via CNAME/DNS.

Not supported

Public preview URLsShareable per-deploy / per-branch URLs.

Not supported

Direct public IPv6Per-instance stable public IPv6 for raw TCP/UDP (beyond a managed HTTPS proxy) — e.g. mosh / raw sockets.

Not supported

Egress controlsConfigurable OUTBOUND network policy — egress firewall / allowlist / egress proxy on traffic LEAVING a workload. Scoped to OUTBOUND only (v15): private-network REACHABILITY between a user's own services is private-networking, not this. Static-egress-IP belongs here (an outbound concept).

Partially supported

Private networkingA platform-internal private network between a user's OWN services — service-to-service connectivity, VPC, service mesh, and/or service discovery, not exposed to the public internet. DISTINCT from egress-controls (OUTBOUND firewall/allowlist), reverse-proxy (PUBLIC inbound ingress), and direct-ipv6 (per-instance PUBLIC addressing). Admitted v15.

Supported

Email send / receiveSend and/or receive email from the workload.

Not supported

Edge KV storeA globally-replicated key-value store read from WITHIN an edge runtime at the PoP (low-latency, eventually-consistent) — Cloudflare Workers KV, Deno KV, Fastly KV Store, Akamai EdgeKV, Vercel Edge Config. Distinct from first-party-db (a managed DATABASE service you provision and connect to) and from http-query-api (HTTP access to an external store) — edge-kv is colocated with the edge compute and read at the edge.

Not supported

Developer access 5/10

Public APIDocumented programmatic API.

Supported

OpenAPI specMachine-readable OpenAPI / Swagger spec published.

Supported

CLIOfficial command-line tool.

Supported

SDKOfficial client library/libraries.

Supported

SSH accessDirect SSH into workloads.

Partially supported

Filesystem APIRemote file operations over the API (list/read/write/watch/copy/upload) without SSH.

Supported

Git deployDeploy from a Git repo / push-to-deploy.

Not supported

Private container registryHosted private image registry.

Not supported

E2B-compatible APIDrop-in compatibility with the E2B sandbox API/SDKs — existing E2B client code runs against this product by changing the endpoint/domain + key, with no business-code changes. E2B itself passes as the reference implementation. Distinct from public-api / sdk (the product has its own API — this key says it also speaks another vendor's) and from self-host-oss (a self-hosted E2B-compatible server is scored on both).

Partially supported

HTTP data APIQuery / read / write the stored DATA over plain HTTP/REST (or GraphQL) WITHOUT the native protocol driver (Postgres wire / Redis RESP) — agent- and edge-friendly data access via fetch(). The DATA plane; distinct from public-api (a control-plane / management API). Auto-generated-from-schema REST/GraphQL (PostgREST, pg_graphql) is a flavour of this.

Not supported

AI & agents 3/11

Built-in coding agentShips a first-party AI coding agent in the product.

Not supported

Agent skillsInstallable / declarable agent skills or tools.

Supported

AI app builderPrompt-to-app / vibe-coding generation surface.

Not supported

MCP serverExposes a Model Context Protocol server for agents.

Supported

MCP clientActs as an MCP client / consumes MCP tools.

Not supported

MCP-server hostingRun/host MCP servers as a workload on the platform (distinct from exposing a first-party MCP control server — that is mcp-server).

Supported

LLM gatewayRoutes LLM API calls through the platform (multi-provider).

Not supported

BYOK for LLMsBring-your-own LLM provider keys.

Not supported

Model inference API (first-party)A FIRST-PARTY hosted LLM/model inference API — the provider serves its OWN catalogue of (usually open-weight) models behind a (typically OpenAI-compatible) endpoint, billed per token. Distinct from llm-gateway (a multi-PROVIDER proxy/router for YOUR calls to external providers like OpenAI/Anthropic) and from byok-llm (bring your own provider keys) — here the platform IS the inference provider, serving its own models.

Not applicable

Autonomous QA agentA first-party AUTONOMOUS AI agent that QA-TESTS a running app / implementation — exercising it like a real user (clicking through flows, checking behaviour) and reporting bugs/defects (often back to a coding agent, or before a human review). Distinct from built-in-agent (a code-AUTHORING agent — this one VERIFIES the result, it does not write the feature; a product can ship both), from ai-guided-code-review (organises a diff for a HUMAN to review — not an autonomous tester), and from observability (runtime metrics/logs, not active testing).

Not supported

Credential-injection proxyThe platform holds real secrets OUTSIDE the workload and injects them on the wire (an egress / broker proxy swaps a placeholder or phantom token for the real credential on matching outbound requests), so code running in the sandbox — e.g. an untrusted agent — never sees or can exfiltrate the real key. Distinct from secrets-management (storage + injection INTO the workload as env vars/files — the workload holds the secret) and from egress-controls (outbound allow/deny POLICY; a product can have an allowlist without injecting credentials, and vice versa).

Not supported

Preinstalled coding agentThe platform installs and launches a THIRD-PARTY coding agent (a harnesses.yml entry — Claude Code, Codex, OpenCode, Pi ...) inside a workspace/sandbox it provisions, with the agent's auth wired in (your own account or API key), so you start an agent session rather than an empty box. Distinct from built-in-agent (a FIRST-PARTY agent the vendor ships) and from harnesses[] (which records that a harness CAN target the product, not that the product runs it for you).

Partially supported

Data 0/4

First-party databasePlatform-provided managed database.

Not supported

Database branchingCopy-on-write branch of a DATABASE (schema + data) as a dev-lifecycle primitive — branch your data like code, usually instant via a metadata operation. Distinct from snapshots-forks (forks a running compute INSTANCE / VM state) and from preview-environments (a full per-PR environment that may USE a DB branch). For a data product a CoW branch maps HERE, not to snapshots-forks.

Not supported

Vector searchFirst-class storage, indexing, and similarity search of vector embeddings — a native vector type/index or a managed vector DB. The data-layer primitive for RAG / semantic retrieval.

Not supported

Ephemeral instant DB (no-signup)Provision a working, usable instance via a single command / HTTP call with NO account, signup, or card — typically time-boxed (expires unless claimed). A zero-onboarding path an agent can hit directly.

Not supported

Platform & operations 0/5

Scheduled tasksCron / scheduled job primitive.

Not supported

Durable workflowsDurable execution / long-running workflow primitive.

Not supported

Preview environmentsA full EPHEMERAL environment spun up per PR / branch — compute + addons + (often) a DB branch — torn down on merge. More than preview-urls (which is just a shareable URL for one deploy).

Not supported

Built-in observabilityFirst-party metrics / logs / tracing / APM / profiling included in the platform (no third-party stack required to see what a workload is doing).

Partially supported

Managed CI/CD runnersPlatform-hosted, drop-in CI/CD runner VMs (GitHub Actions / GitLab), typically faster and cheaper than the CI vendor's own hosted runners, billed per runner-minute; isolated + ephemeral per job. Distinct from git-deploy (push-to-deploy of an APP) and scheduled-tasks (cron) — this is managed CI COMPUTE you point your existing pipelines at.

Not applicable

Project template marketplaceA marketplace / gallery of prebuilt PROJECT STARTER templates (apps or websites) a user clones or launches a new project from and then customises — the "start from a template" surface of a builder / PaaS. Distinct from sandbox-templates (a COMPUTE base IMAGE you fork runtime INSTANCES from — Dockerfile / OCI), from app-builder (the prompt-to-app GENERATION surface — a product can have both), and from one-click OSS-app catalogues / helm-addon-marketplace (deploy a FINISHED app as-is, not a starter you build on).

Not supported

Organisation & security 1/7

Built-in authPlatform-provided authentication / IAM for workloads.

Not supported

SSO / SAMLEnterprise single sign-on.

Not supported

Secrets managementFirst-class secrets storage / injection.

Partially supported

TeamsMulti-user team accounts.

Not supported

SharingShare a running workload/resource via link (optionally gated).

Not supported

BYOCRuns inside the customer's own cloud account.

Not applicable

Deploy into customers' VPCsA SaaS vendor uses the platform to deploy and operate ITS software inside ITS CUSTOMERS' cloud accounts / VPCs (nested or "customer BYOC"), managed from the vendor's control plane. Distinct from byoc (your own workloads in your own cloud account).

Not applicable

Workload identityThe platform gives each workload its own verifiable identity (OIDC token, SPIFFE SVID, cloud IAM federation) that it exchanges for short-lived credentials, so no long-lived static secret is stored or injected. Distinct from secrets-management (stores / injects static secrets) and credential-injection-proxy (swaps real secrets in on the wire).

Partially supported

Self-host / OSS coreOpen-source core that can be self-hosted.

Supported

03 Agent readiness Pre-login surface · counts, not a grade

Can an agent read it?

parse-ability · docs, metadata, machine-readable surfaces

7/10

5 full · 2 partial · 70% of applicable checks

llms.txt
Not supported
“https://agent-sandbox.sigs.k8s.io/llms.txt and /docs/llms.txt: 404.”
llms-full.txt
Not supported
“https://agent-sandbox.sigs.k8s.io/llms-full.txt and /docs/llms-full.txt: 404.”
robots.txt AI policy
Partially supported
“https://agent-sandbox.sigs.k8s.io/robots.txt: User-agent: *; no AI-specific policy.”
JSON-LD
Partially supported
“https://agent-sandbox.sigs.k8s.io/ : WebSite JSON-LD only; invalid zero-year dates in template, no Product offers.”
Sitemap
Supported
“https://agent-sandbox.sigs.k8s.io/sitemap.xml: non-trivial URL set.”
Meta tags
Supported
“https://agent-sandbox.sigs.k8s.io/ : descriptive title, description and OG metadata.”
Markdown negotiation
Not supported
“Accept: text/markdown returned text/html; /index.md 404. Raw GitHub Markdown exists but is not site content negotiation.”
Content efficiency
Supported
“Homepage readable text / HTML bytes = 14.8%; script/style/tag stripped.”
Docs discoverable
Supported
“https://agent-sandbox.sigs.k8s.io/ : first-class Docs navigation.”
Machine-readable pricing
Not applicable
“No first-party hosted tier. Apache-2.0 license, infrastructure costs external.”
No anti-bot wall
Supported
“Homepage, docs and repository sources fetched without public challenge.”

Can an agent drive it?

act-ability · APIs, SDKs, MCP, credentials, webhooks

8/9

7 full · 1 partial · 89% of applicable checks

Public API
Supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/docs/api.md : documented CRDs and sandboxd runtime contracts.”
OpenAPI spec
Supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/packages/sandboxd/spec/filesystem/v1/filesystem.yaml : fetched valid OpenAPI 3.0.3 document.”
MCP server
Supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/clients/integrations/mcp-server/README.md : documented provisioning/command/file tools.”
SDK packages
Supported
“Official Go and Python SDK documentation captured; TypeScript release evidence.”
Auth documented
Supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/packages/sandboxd/USER_GUIDE.md : Kubernetes-authorized port forwarding, no daemon TLS/auth; explicit NetworkPolicy/mTLS proxy guidance; Python SDK gateway auth documented.”
Self-serve credentials
Not applicable
“No hosted account/API key issuance; operator supplies Kubernetes credentials.”
Webhooks
Not supported
“No documented product webhook delivery/signing found in reviewed API/runtime documentation.”
CLI tool
Supported
“README documents official kubectl interface for CRD management.”
Error model
Supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/packages/sandboxd/spec/filesystem/v1/filesystem.yaml : Error schema; runtime guide specifies standard gRPC errors.”
Rate limits documented
Not applicable
“No vendor-operated endpoint or universal hosted rate limits; operator supplies ingress controls.”
Agent config files
Partially supported
“https://raw.githubusercontent.com/kubernetes-sigs/agent-sandbox/main/clients/integrations/mcp-server/skill/README.md : installable MCP skill; no second captured user-facing config surface.”

04 Coding harnesses Which agents and IDEs are documented to run here

Advertised integrations 4

Claude Code Anthropic
Supported
“Any MCP-speaking agent (Claude Code, Cursor, OpenCode) works with zero client library.”
Source raw.githubusercontent.com
Advertised optional integration; not exercised in this Phase-1 capture.
Cursor Anysphere
Supported
“Any MCP-speaking agent (Claude Code, Cursor, OpenCode) works with zero client library.”
Source raw.githubusercontent.com
Advertised optional integration; not exercised in this Phase-1 capture.
OpenCode opencode
Supported
“Any MCP-speaking agent (Claude Code, Cursor, OpenCode) works with zero client library.”
Source raw.githubusercontent.com
Advertised optional integration; not exercised in this Phase-1 capture.
Openhands
Supported
“A workspace backend for the [OpenHands agent SDK]”
Source raw.githubusercontent.com
Advertised integration; registry admission proposed, registry unchanged.

Plausible compatibility 0

Derived, not stored: the product passes SSH access, so any CLI coding agent is installable inside it. This is plausible compatibility, not a supported integration.

No general shell access documented, so nothing is derived. Absence is not unsupported.

05 Compliance & regions Standards held, with ordinal rigor; never a score

Certifications held 0

No product-held standards found in README, threat model or public documentation. Kubernetes affiliation and operator/cloud certifications are not product attestations.

Regions —

No region list captured.

06 Open source Apache-2.0 · fetched by machine, not quoted from marketing

Licence & repository

Licence: Apache-2.0OSI
Primary language: Go
Stars: 4,190
Forks: 549
Created: 12 Aug 2025
Archived: No

Activity

Last commit: 8 Oct 2026
Latest release: v1.0.5 · 1 Oct 2026
Issues open / closed: 124 / 335 · 73% closed

Eight sampled releases from August 13 to October 1, 2026: approximately weekly.

Fetched via gh on 8 Oct 2026.

Self-hosting

Methods: kuberneteshelm
You operate: kubernetes

No hosted tier. Operator supplies cluster, storage, CNI, runtime classes and authentication. GKE memory snapshots need GKE Autopilot/gVisor and PodSnapshot CRDs; optional KEDA scaling needs Prometheus.

Documentation signals 7/8

SupportedQuickstart
SupportedAPI reference
SupportedSelf-host guide
SupportedChangelog
SupportedOpenAPI
Not supportedllms.txt
SupportedExamples
SupportedArchitecture doc

07 Company & presence Facts only, not scored

Company

Headquarters: Not disclosed
Founded: —
Team size: Not disclosed
Hiring: Unknown

Activity

Last release: —
Last blog post: —

Presence 2

Links found on the public surface scanned; not an exhaustive search.

08 Sources & method How this entry was produced

Public surface only

Everything here was read from agent-sandbox.sigs.k8s.io's unauthenticated pages, docs, pricing and repository on 8 Oct 2026. Nothing behind a login was used.

Statuses, not scores

Capabilities are pass, partial or fail against a written rubric (features v21). “Not assessed” means the scan predates the key.

Confidence is separate

Each fact carries its own confidence. Low-confidence facts are shown, never hidden, so you can weigh them.

Schema: scan v5
Features registry: v21
Keys tracked today: 61
Assessed on this entry: 61

Something wrong on this page?

Nearby in the atlas

All Sandbox products →